The present invention relates to a client-server application execution method and system which allow clients to use services provided by servers on an ATM (Asynchronous Transfer Mode) network and, more particularly, to a network server redundancy configuration method and system which effectively operate when a server fails.
A system for performing communication between clients and servers by using TCP/IP (Transmission Control Protocol/Internet Protocol) as a protocol for connection between different types of computer systems has recently been proposed. In this case, the "client" is a computer terminal which starts peer-to-peer communication in accordance with an application program, and the "server" is a computer terminal which receives a communication request from the client and provides a required service in accordance with a server program.
In a conventional client-server application execution system in which clients use services provided by servers on a network, only one server on the network provides a specific service for a given client at a given time. If, therefore, the server that is providing the service fails, the client cannot receive the service. In this case, the client uses an alternative server. This server switching operation is performed in accordance with an application program which operates on the client.
In the system having the above network server redundancy configuration, an application program must store the address information and the like of all the servers on the network to switch servers, and must execute a procedure for connecting a new server. For this reason, a switching operation cannot be performed at high speed. Furthermore, since the application program also performs switching processing, the throughput of the program decreases.